Abstract

Acute myeloid leukemia (AML) cells can modify the bone marrow (BM) microenvironment and “educate” it to promote leukemia progression; e.g., leukemia cells induce transcriptional changes in BM mesenchymal stromal cells (BM-MSC) that support growth advantage and chemoresistance of leukemia cells (Jacamo, R et al., Blood 2014). Given the genetic heterogeneity of AML, and the variable responses of individual patients to therapy, a better understanding of how different AML genotypes can influence their BM microenvironment is needed. We inoculated C57BL/6 mice with syngeneic murine leukemia cells harboring one of the following human AML genotypes: AML1-ETO, MLL-ENL, or MLL-ENL/FLT3-ITD (further divided into p53 wt or p53-/-) (Zuber, J et al., Genes Dev 2009). Leukemia engraftment was confirmed by live imaging (IVIS) and allowed to progress for 2 to 4 weeks, according to leukemia burden. Multiple mice from each condition were sacrificed to allow isolation of sufficient BM-MSC by flow cytometry based on cell surface markers: CD105+, Sca1+, CD106+, PDGF-Rα+ (CD140), and CD45-. Gene expression profiling with Illumina mouse WGv2 BeadArrays was performed on duplicate pools of mice to identify differences between BM-MSC from various AML-bearing mice and BM-MSC from control mice.Heat mapping of highly-variant genes showed that samples clustered according to AML genotype, confirming data quality and suggesting that changes induced in BM-MSC differ with AML genotype: the two BM-MSC samples derived from the MLL-ENL/FLT3-ITD leukemia-bearing mice were nearest neighbors, and also clustered next to the sample from mice injected with MLL-ENL alone. 942 genes were upregulated in the AML1-ETO BM-MSC samples at least 2-fold as compared to control BM-MSC and those from other AML genotypes. The most highly upregulated genes among these included Tenm4 (an inhibitor of chondrogenesis also known as ODZ4) and Plxnd1, a regulator of angiogenesis. For both MLL-ENL/FLT3-ITD samples, 632 genes were upregulated at least 2-fold as compared to control MSC and other AML genotypes, notably including Cited4 (a repressor of HIF-1α transactivation), Bmp1, and Spon2(a TGF-beta-regulated extracellular matrix (ECM) protein upregulated in multiple types of cancer). Ingenuity Pathway Analysis (IPA) implicated upregulation of genes involved in aryl hydrocarbon receptor signaling, a potential inhibitor of HIF-1α by competition for HIF-1β. For genes differing between BM-MSC exposed to p53 WT and -/- forms of MLL-ENL/FLT3-ITD leukemia, IPA suggested that p53 status differentially affects the BM-stroma transcriptome through effects possibly mediated by CFS3, IL-3 and TGF-β1.We also observed sets of genes that were consistently up- or downregulated in BM-MSC by all four AML genotypes. At a minimal 2-fold change, 57 BM-MSC genes were upregulated in all 4 samples, and 60 genes downregulated, by exposure to AML in vivo. Notable among the upregulated genes were Ctgf (12 and 33 fold increase [FI] by 2 array probes), genes related to complement (C4a, C4b, Serping1), Igfbp5 (FI = ~12 by 3 probes), Mgp (matrix Gla protein, an ECM regulator of multiple types of mesenchymal differentiation, FI = ~8) and Cxcl12 (FI = ~6). The latter, the ligand for CXCR4, is expressed by a subset of BM stromal cells and is essential for the maintenance of normal hematopoietic precursors. Gene Set Enrichment Analysis (GSEA) implicated upregulation of genes involved in ECM receptor interaction and cancer stem cells. Common downregulated genes included Ltf, a tumor suppressor and inhibitor of AKT signaling, and PDCD4, a suppressor of tumor-permissive stromal changes downregulated by miR-21, consistent with our finding of increased mir-21 in AML-derived BM-MSC compared to normal donor-derived BM-MSC.In summary, these results support the hypothesis that AML regulates the transcriptomes of BM-MSC, in both shared and genotype-specific ways. Further analysis of the data is ongoing. DisclosuresNo relevant conflicts of interest to declare.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call